Various flat panel display systems have been developed over the last several decades. Among them is the Time Multiplexed Optical Shutter disclosed in Selbrede U.S. Pat. No. 5,319,491 (which is incorporated in its entirety herein) and such variations as have been subsequently filed in commonly owned U.S. Pat. Nos. 7,042,618, 7,057,790, 7,218,437, 7,486,854 and U.S. Patent Publication No. 2008/0075414. The fundamental premise of such devices is that light (usually monochromatic light) is edge-injected into a transparent rectangular slab waveguide such that total internal reflection (TIR) of the injected light obtains within the waveguide, which may be mirrored on one or more of the side surfaces to insure maximum transits for rays traveling within the waveguide. The principle of operation for any of the plurality of pixels distributed across the slab waveguide involves locally, selectively, and controllably frustrating the total internal reflection of light bound within the waveguide to emit light at that pixel location. In one pixel architecture, frustration of TIR light bound within the waveguide is achieved by propelling (i.e., moving) an optically-suitable material across a microscopic gap, such that the material is at or near contact with a surface of the slab waveguide in the active position, while in the inactive position the material is sufficiently displaced from the surface of the waveguide so that light and/or evanescent coupling across the gap is negligible. The optically-suitable material, herein referred to as an “active layer”, being propelled (i.e., moved) can be an elastically deformable thin sheet (thin layer or film) of polymeric material (e.g., elastomer) with a refractive index selected to optimize the coupling of light during the contact/near-contact events. Switching the active layer between inactive and active positions can occur at very high speeds in order to permit the generation of adequate gray scale levels for multiple primary colored light (e.g., consecutive primary colored lights red-green-blue) at video frame rates in order to avoid excessive motional and color breakup artifacts while preserving smooth video generation. The flat panel display is thus comprised of a plurality of pixels, each pixel representing a discrete subsection of the display that can be individually and selectively controlled in respect to locally propelling the active layer bearing a suitable refractive index across a microscopic gap into contact or near contact with the slab waveguide. The propulsion can be achieved by the electromechanical and/or ponderomotive deformation of the thin sheet of polymeric material, said sheet being tethered at the periphery of the individual pixel geometry by standoffs that maintain the sheet in a suitable spaced-apart relation to the slab waveguide when the pixel is in the quiescent unactuated state. Application of an appropriate electrical potential across a first conductor disposed on or within the slab waveguide and a second conductor disposed on or within the active layer, causes the high-speed motion of the active layer toward the surface of the slab waveguide; actuation is deemed completed when the active layer can move no closer to the slab waveguide (either in itself, or due to physical contact with the waveguide). To facilitate light extraction, an array of micro-optical structures (of various possible geometries, such as frustums or pyramidal sections, etc.) may be optionally disposed on the waveguide-facing side of the active layer, such that pixel actuation entails contact or near-contact of these micro-optical structures with the waveguide, thus frustrating TIR light in such a way that re-direction of extracted light to the viewer is optimized. A more detailed description of micro-optical structures is disclosed in “Optical Microstructures for Light Extraction and Control” U.S. Pat. No. 7,486,854, which is incorporated herein by reference in its entirety.
Certain other display systems use similar (but not identical) principles of operation. Some utilize a backlight system where the pixels literally shutter light, usually by transverse lateral motion of an opaque MEMS-based shuttering element at each pixel parallel to the main surface (e.g., top surface) of the waveguide configured as a true backlight system proper, contra the TIR-based waveguide of Selbrede '491 which is not a true backlight given the TIR-bound condition of light traveling inside it. For a backlight system, light within the slab waveguide should not be maintained in a TIR-compliant state lest it be perpetually bound to the interior of the waveguide. Thus, the bottom surface of the waveguide can be made a scattering surface, or it can diverge from a parallel spaced-apart relation to the top surface of the waveguide, or both, to insure that light continually departs the top surface of the slab waveguide to illuminate the pixel shutter mechanisms arrayed at or above the top surface of the slab waveguide. The appeal of using a slab waveguide for transverse MEMS shutter-based systems is due to the ability to recycle unused light by configuring the waveguide-facing portions of the shutter mechanisms to be nominally reflective. Light not passing through an open shutter may then re-enter the waveguide and can be used elsewhere within the system.
In the case of devices based on Selbrede '491, in which the light sources are arrayed on one edge of the slab waveguide while the opposite end from said edge is mirrored (with either a metallic reflector disposed thereon or by imposition of a perfect dielectric mirror to gain even better reflectance), it has been determined that the luminous uniformity of the display can only be insured when the thickness of the slab waveguide is sufficiently thick. A minimum slab waveguide thickness, t, that can be utilized for the slab waveguide is a function of the length of the waveguide l, the critical angle of the waveguide θc (which is itself a function of the waveguide's refractive index), and the individual optical efficiency of a pixel on the display surface, denoted ∈. The mean free path of a given photon ensemble from origin at the light source to 99% depletion inside the waveguide is given the Greek symbol λ, which is not to be confused with the optical wavelength of that light in this context. By detuning the effective individual pixel efficiency ∈, and using the resulting average mean free path of a photon ensemble prior to 99% depletion, λ, uniformity has been demonstrated to be readily optimized when λ=3l or greater, thereby establishing a lower bound on slab thickness by the following equation:
  t  =            3      -                        (                                    log              ⁡                              (                0.01                )                                                    log              ⁡                              (                                  1                  -                  ɛ                                )                                              )                ⁢        l                    (                                    cos            ⁡                          (                              θ                c                            )                                ⁢                      log            ⁡                          (              0.01              )                                                log          ⁡                      (                          1              -              ɛ                        )                              )      
Applying this constraint to the slab waveguide thickness enables displays based on such waveguides to achieve in excess of 60% optical efficiency (ratio of light flux input to light flux output) while simultaneously insuring far less than 1 dB variation in luminosity across the entire display surface (typically under 0.2 dB variation).
While this constraint is of minimal consequence for many applications, it does present a step backward for applications where the industry trend has been toward thinner display subsystems year after year. Thus, for a cell phone, the thickness constraint might require the waveguide to be up to 2 millimeters thick or more to insure outstanding luminous uniformity, whereas the trend in cell phone display components is for the display to be under 1 mm in total thickness. In actual fact, a waveguide thickness of 0.7 mm is desirable, given that this is a standard thickness for LCD mother glass and TFT active matrix glass. However, so thin a waveguide, by violating the thickness t constraint outlined above, runs a serious risk of suffering from debilitating nonuniformities in brightness across the display surface. The symbol t shall hereafter be denominated the minimum slab waveguide thickness that corresponds to the minimum luminous uniformity threshold limit.
Recent co-pending filings have disclosed various apodization (compensation) means in orienting and configuring the illumination means at the edge(s) of the waveguide (e.g., a varying distribution of light sources along an edge of the waveguide) to resolve luminous nonuniformity. However, the periodicity of the pixels and/or micro-optical structures disposed on the light extraction surface of the display system (e.g., top surface of the slab waveguide), in conjunction with the point source nature of the illumination means (e.g., multiple discrete LEDs), has given rise to other optically undesirable effects, such as Moiré patterns, banding, headlighting (ability to resolve the individual light sources illuminating the display system), and other light artifacts created by using discrete light sources (e.g., LEDs) to feed light to the waveguide. These light artifacts can be sufficiently severe as to create liabilities for displays that otherwise may exhibit reasonable macro-level uniformity. It is an object of the present application to address these artifacts at the illumination source by making the light entering the waveguide sufficiently diffuse (e.g., uniform) that the periodic intensity of the original light from the individual light sources can no longer be individually resolved.